Frederick C. Wilt said:
- The lack of a single login
There's a reason for that, as there often is, but it's not necessarily obvious. Some programs are only available to user of a particular level. You can't use a single login because that would prevent an admin from logging in as a different user from someone who is currently logged in as a lesser rights user. It would also allow someone to walk up to a machine where an admin might be logged into something like the IV, and to just start up the Admin Interface and do anything he wants.
Frederick C. Wilt said:
I know you have your stated reasons but not using a install process that is consistent with the majority of user experiences makes an impression.
That's already been dealt with, as of about a week ago.
Frederick C. Wilt said:
This is where your priorities as a programmer take precedence over good human factors engineering.
But it's really not. It's what's practical given the time available, and sometimes within technical constraints that would take too long to get around given the time available. I know what's good and what's not. I use just as many programs as all of you. But great doesn't come cheap in terms of time. You wouldn't believe the time suckage that a creating something like a complex dialog box takes, even to just do it fairly simply. And there are hundreds in CQC.
And, in some cases, it's because some bits were written a decade ago, and rewriting them would take even more time than just starting over, which is likely what's going to happen (for the Admin Interface at least.) And it's maybe even worse to have some of the same program to work one way and some another, so in that way it's sort of also difficult to do it without starting over. And in some cases to do it better would require changing how the back end works, but that would require that lots of other stuff be changed to follow, again easier if just starting over.
It doesn't require a degree in UI design to do a good job. But it does require a butt load of time. Of course, if I do start over, I'd know a lot more this time around in terms of what is required, and I wouldn't be constrained by existing stuff. And if I do start over, I'd do it incrementally, i.e. the first thing would be some sort of limited function version of the product, so that I'd have a much shorter row to hoe to get something sellable out, could work out the issues in a less complex initial environment, and could leave the existing system alone in the meantime and not break it until I'm ready to take it on.